Quantitative characterization of mouse embryonic stem cell state transition

Abstract

It is known that on-and-off transcription factors (TFs) regulations play crucial roles in differentiation and reprogramming. Oct4, Sox2 and Nanog as well as their constituted gene networks cooperatively maintain ES cell pluripotency. Recently, TFs such as Nanog, Stella and Rex1 show heterogeneous expression in ES cell populations and such heterogeneity lead to different differentiation potential indicating that heterogeneous gene expression may control cell fate. Besides gene regulation triggered by TFs, intrinsic noise was also found to be a potential source in heterogeneity and stochastic cell fate determination. For safe clinical application of stem cell-based therapy, it is crucial to quantify these dynamic processes and transition rate during research and development. However, in most cases, the transition rate is too low to be measured and is often entangled with cell proliferation. In order to disentangle these processes, we established a Nanog-EGFP reporter ES cell line and quantified the transition rates between two subpopulations. We also developed a mathematical model involving cell state transition, proliferation and noise to facilitate deciphering this process. We first generated a dual reporter ES cell line (Nanog-EGFP and EF1α-H2B-mCherry) based on PiggyBac system. Using a simple ES cell differentiation process, differentiation and dedifferentiation could be obviously observed based on Nanog ES cell reporter. In order to quantify the transition and proliferation rate, we performed multiple batches of ES cell differentiation and dedifferentiation based with this reporter ES cell line. Mathematical models (drift-diffusion-growth equation) are also introduced to fit the data. The quantitative and nontrivial model predictions lead to new experiments and hypotheses. Combined the flow cytometry data with mathematical models, we are able to directly derive the most likely potential landscape (Waddington’s landscape) based on Nanog distribution. Our results and approaches developed would provide more clues for future ES cell differentiation studies. In another project, we generated multiple ES cell reporters in which Nanog, Oct4 and Rex1 were all labeled with different fluorescence proteins. We differentiated the reporter ES cell line specifically into epiblast stem cell (EpiSC) based on previous protocols. Based on flow cytometry analysis, we found that, during the transition from ES cell to EpiSC state, all the three marker genes showed heterogeneous expression pattern, and three sub-populations can be identified at the EpiSC steady (later) state. The cell fate identity and potential converting ability into other identities like ES cell or inter-converting ability among three populations were being conducted based on chemical factors. In addition, we quantitatively characterized PiggyBac mediated multiplex gene transfer in mouse embryonic stem cell which lay a good foundation in above mentioned ES cell reporter generation. HS4 insulator was also characterized on the effect of protecting from promoter interference and gene silencing as well as transposition efficiency. Our results suggest that, even after the development of CRISPR and TALEN, transposon technique is still a powerful tool to study synthetic gene circuit in mammalian systems, especially for prototyping and complex circuits with many parts.